Interest in use of bivalves to remediate estuarine eutrophication has increased in recent years. High variation among data sets, substantial focus on particle removal, and insufficient links to anthropogenic nitrogen (N) sources encouraged this empirical examination of N removal by bivalves from estuaries receiving different N loads. We determined the capacity of the oyster Crassostrea virginica to remove N by comparing N assimilated into tissues with anthropogenic N from land or available in phytoplankton. Oyster growth yielded 0.2-0.4 g N in tissues and depended on estuary-specific conditions. d 15 N values confirmed that N in oyster tissues derived from local anthropogenic sources. At representative restoration and aquaculture conditions (≤400 oysters·m -2 at 0.5%-1.0% of estuary area), estimated N removal was ≤15% of landderived loads and <1% of phytoplankton N. N removal via biogeochemical processes was negligible during grow-out, but became important after oysters attained harvestable size. This study explicitly demonstrates that oysters assimilated landderived N, but suggests that bivalve bioremediation consider trade-offs between intensity of planting, ecological effects, and available space.Résumé : Le recours aux bivalves pour contrer l'eutrophisation estuarienne suscite un intérêt croissant ces dernières années. La variabilité élevée d'un ensemble de données à l'autre, un accent important sur le retrait de particules et l'insuffisance de liens établis vers des sources anthropiques d'azote (N) ont motivé l'examen empirique du retrait de l'azote par des bivalves d'estuaires qui reçoivent différentes charges d'azote. Nous avons déterminé la capacité de l'huître, Crassostrea virginica, de retirer l'azote en comparant le N assimilé dans leurs tissus au N anthropique de source terrestre ou disponible dans le phytoplancton. La croissance d'une huître produit de 0,2 g à 0,4 g de N dans les tissus, selon les conditions propres à l'estuaire. Les valeurs de d 15 N ont confirmé que le N dans les tissus d'huîtres provient de sources anthropiques locales. Dans des conditions de restauration et d'aquaculture représentatives (≤400 huîtres·m -2 sur de 0,5 % à 1,0 % de la superficie de l'estuaire), le taux de retrait estimé de l'azote était de ≤15 % des charges de sources terrestres et de <1 % du N provenant du phytoplancton. Le retrait du N par des processus biogéochimiques s'est avéré négligeable durant le grossissement, mais devenait important après que les huîtres aient atteint une taille commerciale. L'étude démontre explicitement que les huîtres ont assimilé du N de source terrestre, mais suggère que la biorestauration à l'aide de bivalves devrait tenir compte des compromis entre l'intensité de l'ensemencement, les effets écologiques et l'espace disponible.[Traduit par la Rédaction]

Data on N assimilation into oyster tissues were compared with N in phytoplankton (Ph N) to calculate the number of oysters required to completely remediate N loads and determine the percent N removed by oysters at different planting densities and areas. After publication, we discovered a conversion error that was made when calculating the Ph N load relative to estuary volume, and we accounted for the relatively high estimates of Ph N we reported compared with values reported elsewhere. Corrected estimated values for Ph N available to oysters based on estuary volume had a range of 44-281 kg N (Table 1), assuming a consistent chlorophyll a concentration without replacement for the 153-day estimated growing season (necessarily a conservative estimate). Under this reduced estimated Ph N load, a minimum of 250 000-770 000 oysters would be required to assimilate Ph N in the estuaries we studied (Table 1), assuming the entire estuary was available and suitable for planting. N removal capacity was estimated at 17%-100% of Ph N (consistent with or greater than the N removal capacity estimated for total land-derived N loads), with total Ph N removal possible in scenarios where oysters were planted at densities ≥550•m −2. Since estimates of Ph N depend heavily on estuary-specific physical and chemical attributes not accounted for in this study, as with our original estimates, we suggest these values be used with caution and estuary-specific empirical data be collected for application to other locations and growing conditions. This correction does not change our total land-derived N load comparisons or the conclusions of our originally published work, that a removal capacity of 1%-15% of land-derived N is most realistic given conditions typical to restoration and aquaculture activities in USA estuaries.